We have previously used various empirical methods to try to derive information from amino acid sequences of proteins and to use such information to predict possible three dimensional foldings of different proteins. Some of these predictions were resonably successful, while some not. It is therefore important to refine our empirical methods to give better predictions. Such refinements may be achieved by a detailed analysis of those proteins whose tertiary structures have been determined by X-ray diffraction studies. Roughly, they can be classified into three different groups: (1) those consisting of mostly alpha-helices, e.g. myoglobin, (2) those consisting of mostly Beta-sheets, e.g. concanavalin A, and (3) those consisting of some alpha-helices and some Beta-sheets, e.g. lactate dehydrogenase. If we want to predict, for example, the structure of a protein which contains essentially Beta-sheets, e.g. any short or long snake neurotoxin, only those in group (2) should be used as reference proteins. In addition, we would like to study two specific problems: (a) The hypervariable segments of immunoglobulins - Since the non-hypervariable segments of the variable regions of light and heavy chains of immunoglobulins are structural and form essentially an invariant three dimensional structure, we can use that as a fixed structure and try to predict the tertiary foldings of the hypervariable segments so that they can fit into that framework. (b) Ribitol dehydrogenase - More than half of the molecule is used to bind the coenzyme, nicotinamide adenine dinucleotide, and that portion may have a similar tertiary structure of those of alcohol, lactate and glyceraldehye 3-phosphate dehydrogenases. If this is true, it is then only required to try to predict the folding of a small portion of the molecule.